[Gate-users] Potential bug in units to materials conversion

Matthew Strugari matthew.strugari at dal.ca
Tue Mar 3 22:06:58 CET 2020


Hi David(s)!

Sorry for the incorrect terminology, bare with me :) I was expecting to see the dose in air fall off at a faster rate when moving away from the kidney, especially considering that the CSDA range for Y-90 betas is ~1000x larger in air than in water when taking density into consideration. Therefore, I would expect the absorbed dose to decrease significantly in air and become more spread out. Rather, it looks like the dose continues to fall off beyond the "air-water interface" as if the electrons were continuing in water and not air. I simulated a quick test of a dose point kernel in air vs water and there are subtle differences, but I guess the differences are not as drastic as I expected. However, with your explanation of how "Air" lands in the materials database, I will remain confident that material properties are being properly assigned and there is no "bug" in the materials conversion.

I am using a 2.22 um (1 keV) range cut in water with the step size set to default due to my 1.5 mm isotropic voxel volume; my stepHitType is random. Unfortunately, tracking only shows interactions for phantom_voxel_phys_Z with no associated material information so I cannot tell in which material transportation occurs.

My apologies for my naive use of the function, I assumed it would build an image from the provided filename and dump the corresponding info. I will look closer into the functions next time. Nonetheless, the input file was not overwritten and providing a new filename did not produce output.

Thank you for taking the time to respond!
Matthew


________________________________
From: David Boersma <david.boersma at acmit.at>
Sent: March 2, 2020 8:24 PM
To: Matthew Strugari <matthew.strugari at dal.ca>; David.Sarrut at creatis.insa-lyon.fr <David.Sarrut at creatis.insa-lyon.fr>
Cc: gate-users at lists.opengatecollaboration.org <gate-users at lists.opengatecollaboration.org>
Subject: AW: [Gate-users] Potential bug in units to materials conversion


Hi!



Sorry for my ignorance: what are “backscatter effects at the water-air interface", how should those effects look like in the VV images that you sent in the first mail in this thread? What physics cuts (range, energy, stepsize) are you using, and what is the configuration of "stepHitType" in the dose actor (mid, pre, post, random)?

To me it looks like you provide "buildAndDumpLabeledImage" with the same (output) filename as the input image, so maybe you don't see any output because instead of producing a new file it overwrote the input file?

In the first mail in this thread, the text output about the three materials ("worldDefaultAir", "Air" and "Water") seems to be printed by the main "GateDetectorConstruction" class,  not by the class that directly manages voxelized images. This printout is produced by looping over "the material database", which is a dictionary-like object that contains all materials that have been defined or looked up during the simulation run (and the input for that table might come from the collection of Geant4/NIST materials and/or from text files such as "GateMaterials.db"). Some grepping in the code shows that the constructor of the base class for image volumes (including the ImageNestedParametrisedVolume class) initially sets the material of the volume to "Air", and by doing that, "Air" lands in the material database. Afterwards the derived class may define materials different from "Air" in the entire volume ("Water" and "worldDefaultAir" in your case), but once "Air" is in the database it stays there until the simulation run ends, even if it's not actually used anymore.

(Now why are "Air" and "worldDefaultAir" actually referring to different materials, with e.g. slightly different ionization potential, instead of just two different names for the same materials? That's a topic for a new discussion. This mail was supposed to be a short response and it got way longer than I intended...)


HTH,
David B.

________________________________
Von: Gate-users <gate-users-bounces at lists.opengatecollaboration.org> im Auftrag von Matthew Strugari <matthew.strugari at dal.ca>
Gesendet: Montag, 2. März 2020 18:54:45
An: David Boersma; David.Sarrut at creatis.insa-lyon.fr
Cc: gate-users at lists.opengatecollaboration.org
Betreff: Re: [Gate-users] Potential bug in units to materials conversion

Hi David,

No luck with "buildAndDumpLabeledImage". I implemented as
/gate/phantom/geometry/buildAndDumpLabeledImage             data/Cropped_kidneyContour_binary.mhd
but there was no associated output.

I assume this is the correct usage as suggested by the --qt viewer.

Regards,
Matthew

________________________________
From: David Sarrut <David.Sarrut at creatis.insa-lyon.fr>
Sent: March 2, 2020 11:56 AM
To: Matthew Strugari <matthew.strugari at dal.ca>; David Boersma <david.boersma at medaustron.at>
Subject: Re: [Gate-users] Potential bug in units to materials conversion

Hi Matthew,

quick idea: try "buildAndDumpLabeledImage" to see what happens ...

DavidB : any idea ?

Thanks,
David

On Mon, Mar 2, 2020 at 4:33 PM Matthew Strugari <matthew.strugari at dal.ca<mailto:matthew.strugari at dal.ca>> wrote:
Hi everyone,

As a followup to my previous email, it appears that the bug affects voxelized simulation results. After allowing the simulation to run to completion, the dose distribution due to 40 million Y-90 primaries does not show any back-scatter effects at the air-water interface. I have attached an image which highlights a couple regions showing the extension of dose into "air". The dose decreases smoothly with distance from the source despite crossing the boundary.

I use the following lines to define my voxelized phantom:
/gate/world/daughters/name                                  phantom
/gate/world/daughters/insert                                ImageNestedParametrisedVolume
/gate/phantom/geometry/setImage                             data/Cropped_kidneyContour_binary.mhd
/gate/phantom/geometry/TranslateTheImageAtThisIsoCenter     0 0 0 mm
/gate/phantom/geometry/setRangeToMaterialFile               data/KidneyAttenuationRange.dat

I have also changed the attenuation range to overlap with values near 0 and 1 in case of any floating point errors to no avail. If anyone sees any mistakes in my phantom definition or has encountered similar issues, please let me know. I am not sure what to do about this.

Regards,
Matthew

________________________________
From: Gate-users <gate-users-bounces at lists.opengatecollaboration.org<mailto:gate-users-bounces at lists.opengatecollaboration.org>> on behalf of Matthew Strugari <matthew.strugari at dal.ca<mailto:matthew.strugari at dal.ca>>
Sent: February 28, 2020 12:43 PM
To: gate-users at lists.opengatecollaboration.org<mailto:gate-users at lists.opengatecollaboration.org> <gate-users at lists.opengatecollaboration.org<mailto:gate-users at lists.opengatecollaboration.org>>
Subject: [Gate-users] Potential bug in units to materials conversion

Dear gate-users,

There appears to be an issue when defining the associated material for each voxel in an image. I have loaded a binary image with the units to materials conversion defined in a AttenuationRange.dat file:
2
0 0 worldDefaultAir
1 1 Water
I have also set the world material as worldDefaultAir resulting in a total of two defined materials throughout my entire simulation. When executing my macro, there is a message stating that 3 materials are in the image:
[Volume-1]  Number of different materials in the image data/Cropped_contour_binary.mhd : 3
[Physic-1]
[Physic-1]  ----------------------------------------------------------
[Physic-1]                      Ionization potential
[Physic-1]   - worldDefaultAir defaut value: I = 85.7031 eV
[Physic-1]   - Air defaut value: I = 85.6694 eV
[Physic-1]   - Water defaut value: I = 68.9984 eV
[Physic-1]  ----------------------------------------------------------
Has anyone else noticed this issue? I don't know if it's possible to verify what voxels are assigned to what material. It would be nice to see the total number of voxels assigned to a given material at run time.

Best,

Matthew


--

Matthew Strugari

Biomedical Translational Imaging Centre - BIOTIC,
5890 University Ave,
Halifax, NS, B3K 6R8

_______________________________________________
Gate-users mailing list
Gate-users at lists.opengatecollaboration.org<mailto:Gate-users at lists.opengatecollaboration.org>
http://lists.opengatecollaboration.org/mailman/listinfo/gate-users


--
David Sarrut, Phd
Directeur de recherche CNRS
CREATIS, UMR CNRS 5220, Inserm U1206
Centre de lutte contre le cancer Léon Bérard
28 rue Laënnec, 69373 Lyon cedex 08
Tel : 04 78 78 51 51 / 06 74 72 05 42
http://www.creatis.insa-lyon.fr/~dsarrut
_________________________________
 "2 + 2 = 5,  for extremely large values of 2"
_________________________________
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opengatecollaboration.org/pipermail/gate-users/attachments/20200303/165c6f9c/attachment-0001.html>


More information about the Gate-users mailing list